be little doubt that such mosques represent the most sustained attempt in all of Islamic architecture to reconcile the divergent aims of royal and religious iconography. Planted as they are like lances in the sacred precinct, these minarets may be said to consummate a tradition stretching far back in time to the Prophet s long spear ^anaya) which he would thrust into the.ground to indicate the direction of prayer. The wheel has turned full circle and the original form has become splendidly monumentalised. Whether or not this was a deliberate echo is an open question. At all events, it is hard to overlook the aggressive and ceremonial implications of these mosque - in the case just cited, for instance, the outer minarets flanking the principal fa ade of the building are shorter than those flanking the dome. Thus a pyramidal effect is achieved which is still further emphasised by the choice of a sloping site. The gently rolling skyline of Istanbul, with its extensive natural views, was ideally suited to this kind of display, and the political significance of the city as the Ottoman capital may partly have motivated this new use of the minaret as a component of urban design on a mammoth scale. Such minarets were also used in a more symbolic way as markers of the courtyard, of the sanctuary, or of the entire mosque, staking out the boundaries of the religious domain within a secular environment. Dome chamber and minaret alike thus acquire extra significance as symbols of the faith. This development was not new, but only in Ottoman architecture is it pursued with such singlemindedness. It is therefore entirely appropriate that these minarets, like the domes over the mihrab, should bear the emblem of the crescent, supported on a series of superposed orbs. 2.354 3.70 87, 3.71 Luxor, Mosque o f Abu l-hajj, minaret gigantic needle-sharp lances clustered protectively, like a guard of honour, around the royal dome. Their impact depends to a large extent on their proportions, which are almost unprecedented; the pair of minarets flanking the Siileymaniye dome are each some seventy metres high. Such minarets function simultaneously to enrich the exterior silhouette of the MINARETS IN EGYPT If conservatism may be termed the hallmark of the Ottoman minaret, its Egyptian counterpart is above all varied. This variety is all the more remarkable because the Egyptian school is to all intents and purposes concentrated on the buildings of Cairo, though it is represented in some small measure in the provincial towns of Egypt and in the architecture of the Mamluks in Syria and the Levant. Unfortunately very few surviving pre-mamluk minarets have escaped extensive alteration. Moreover, the most important examples to fall within this category are not metropolitan work at all but are found in various provincial towns: Isna, Luxor, Aswan and nearby Shellal. All date from the late 11th century. They already display the characteristic Egyptian division of the minaret into separately conceived superposed tiers. The Asna minaret (474/1081-2) illustrates the type in its classic form. From a square base some thirty-five feet high, generously articulated by windows, rises a plain tapering truncated cylinder capped by an open pavilion whose eight concave sides bear a diminutive hexagonal domed aedicule, also of open plan. Inside the structure is a square newel 3.116, 3.117, 3.121 117, 3.115 3.115 165
ISLAMIC ARCHITECTURE 3.25, 3.59-3.6 2. 3.63 3.127-3.128 2.90, 2.94 staircase with a series of short, sharp ascents. Other minarets of this group maintain the threefold division of the elevation but change the proportions (for example reducing the crowning pavilion, as at Luxor), the decoration or the material (thus the Luxor minaret is of mud brick). Their material, and certain structural features, such as the lantern on free-standing columns and the tapered cylindrical shaft above a lofty square base, have been persuasively linked by Jonathan Bloom to contemporary architecture just across the Red Sea in the Hijaz. Yemeni minarets perpetuate some of these features. Interesting as these minarets are stylistically, they are insignificant in comparison with the great corner towers marking the main fa ade of the mosque of al-hakim in Cairo, built between 380/990 and 401/1010. With their massive, embattled but later square bases, whose taper, like that of an ancient Egyptian pylon, is so pronounced that it is almost a slope, they have all the appearance of bastions. In its original layout the Hakim mosque maintained a powerful consonance between minarets and portal. Very soon, however by 401/1010 each minaret was enclosed by a huge salient some 1.7 m. square, which allotted it a portentous, indeed revolutionary, role. Finally, in 480/ 1087, Badr al-jamali enlarged the northern salient to gigantic proportions (some 25 m. square) and thereby gave that minaret a military function. In so doing he also incorporated the principal fa ade of the mosque into the expanded fortifications of the city and gave it a quasi-military aspect; but he managed to make the minarets play a major part in this process without noticeable strain or incongruity. Even so, it must be admitted that the bastions constitute brutal, unadorned masses of masonry; the minaret shafts above are not only dwarfed by the bulk of their substructure, but also by contrast loaded with architectural and applied ornament. The northern minaret observes the multiple division of parts so typical of the Egyptian style. Its lowest part is a cylinder resting on a cube. Then comes an octagonal shaft with a blind arch and windows on each side, which gives way to a heavy band of muqarnas decoration in three distinct tiers. A fluted keel-shaped dome crowns the whole; within is a spiral staircase. In the western minaret the octagonal muqarnas zone is reduced in size and the square lower shaft is pierced by a double tier of arched windows. But its ornament, featuring two bands of epigraphy and two of arabesque, with numerous additional geometrical panels and cartouches, is significantly richer. Since the minarets of the Hakim mosque survive in such an altered state, it is not easy to see where they belong in the corpus of Egyptian 3.114 3.59 3.60 San a, Masjid al-abhar San a, Great Mosque Sa da, Masjid al-shamri 166 3.62 Sa da, Masjid Ulayyan
3.112 118 Cairo, Mosque o f al-hakim, northern salient and minaret, from the west minarets. This is all the more serious a lacuna in view of the once-vigorous controversy over the role of the Pharos of Alexandria, which stood well-nigh intact until it was partially ruined by an earthquake in 180/796-7, in the evolution of the Egyptian minaret. Pace Creswell, who argued against any connection between the two building types, it can scarcely be overlooked that the surviving Egyptian minarets which date before 493/1100 all attest a pronounced multipartite division of the elevation. Since this feature - though present in the minarets of Mecca and Medina, very possibly as a result of Egyptian influence - is absent alike in the Syrian, Iranian and Maghribi traditions (with two significant exceptions), some rationale for such an unusual division must be proposed. Interestingly enough, the two major Maghribi minarets with three superposed stories are those of Qairawan and Sfax. In the early Islamic period these sites were the first major Muslim settlements on the road west from Alexandria. Moreover, it was precisely in Tunisia, a maritime frontier area in the war against the Byzantines in southern Italy and Sicily, that the building of lighthouses is copiously recorded in the early Muslim sources. As noted above, the form of the Qairawan minaret has itself recently been linked with that of a Roman lighthouse nearby. Thus the idea of an association between lighthouses and minarets, which so mesmerised scholars earlier this century, has not entirely lost its relevance. Quite aside from this, the reasons adduced by Creswell for rejecting any link between the Pharos and Egyptian minarets are themselves not entirely sound. His narrowly chronological approach is superficially attractive because of its methodological rigour. Yet not all types of architectural evolution are entirely chronological. The case of the Holy Sepulchre indicates that the idea of a seminal building may find extremely varied expression at the hands of subsequent architects, and that references to it include copies both very faithful and very distant. CreswelPs proposed evolution effectively ignores the likelihood that a monument as world-famous and as physically memorable as the Pharos would have exerted a continuing influence on Egyptian architecture long after its destruction. If the Pharos can be proposed - though with all due reserve - as a possible source for certain three-staged minarets outside Egypt, its influence within that country is still more likely. This is not to say that any surviving Egyptian minaret is intended even as a reasonably close copy of the Pharos. Instead they might well be regarded as very free variations on the Pharos theme. The principal points of contact would then be the multiple (usually triple) division of the elevation, with superposed storeys of successively reduced diameter and size, and the provision of a crowning open-plan lantern. In conclusion, it is perhaps worth remembering that the Pharos was repeatedly rebuilt by the Muslims until its final disappearance some time between the early 13th and the mid-14th 3.29, 3.30 167
ISLAMIC ARCHITECTURE tier example adjoining the mosque of al-juyushi (478/1085), does it follow that this influence was continuous? The evidence of a host of minarets beginning with that of the mausoleum of Abu l- 5.175-5.176 119 Cairo, minaret of Bashtak century. Indeed, as Butler noted, the account of Abd al-latif indicates that in c. 597/1200 the Pharos comprised successively square, octagonal and round storeys and was crowned by a lantern or small cupola. It may well be, therefore, that this semi-islamic Pharos rather than the original building was the means of establishing the tradition of the multi-staged minaret in Egypt. If, then, it is possible that the Pharos, whether in its original guise or in one of its later transformations, exerted decisive influence on at least some early Egyptian minarets, such as the four- 168 120 Cairo, Ghanim al-bahlawan mosque, minaret
121 Cairo, Mughalbay Taz mosque, base of minaret 169
ISLAMIC ARCHITECTURE 5.176 5.213, 3.122 4.56-4.57 Ghadanfar (552/1157) suggests that this is not the case. In the early versions of such towers the emphasis is on a tall square shaft of Syrian type, which may be very plain (mausolea of Abu l- Ghadanfar and Fatima Khatun) or richly decorated (minaret in madrasa of Sultan al-nasir Muhammad). Crowning this shaft is the socalled mabkhara, a two-storey octagonal pavilion whose dome above a heavy muqarnas cornice is usually fluted and whose lower walls are broken by decoratively profiled arches (examples attached to the %awiya of al-hunad and the madrasa of Sultan Salih). Such buildings, which are mostly of the 13th century, do seem to be independent of the Pharos tradition. It is with their immediate successors that the problem becomes acute. Now the mabkhara is accorded much more emphasis than hitherto, with a consequent downgrading of the main shaft, and the internal divisions of the mabkhara are much more marked. In effect it becomes two separate storeys, whose formal and decorative independence from each other is underlined by the use of different ground-plans: an octagonal storey giving way to a circular one which bears the crowning dome and finial. Thus the Pharos pattern - of tiers which are in turn square, octagonal and circular and are capped by a roof with a crowning device reappears. But does it issue from the Pharos itself, via such transitional monuments as the Juyushi minaret, or is it a natural development of the Abu l- Ghadanfar type? The minaret attached to the Sultan Qala un complex suggests the first alternative, while the almost contemporary minaret of the /tftf /rajv/-mausoleum of Salar and San jar al-jauli suggests the second. Yet for all that, the differences between them are slight. If these were indeed two separate strands in the evolution of the Egyptian minaret, these strands fused in the early 14th century in the minarets of the madrasa-cum-mansoleum of the Amir Sunqur Sa٤di or of the khanqah of the Amir Qusun. By that time (735/1335) the tripartite division was standard. The principle of altering the ratio of one tier vis-à-vis the other continued in later Cairene minarets. Its most striking expression may be seen in the continued reduction of the main shaft, which finally diminishes to the point where it is lost in the surrounding walls of the mosque. Thus the visible part of the minaret is an octagonal shaft with a cylindrical superstructure (minarets of Shaikhun and Sarghitmish, both of the mid-14th century). The future course of the Egyptian minaret was now clear. With the rejection of the tall square shaft as the essential defining feature of the minaret, the way was open for quite radical changes in the proportional relationships between the various parts of the minaret (minaret of khanqah of Faraj b. Barquq). Sometimes the elevation was dominated by a series of diminishing octagons. Multiple balconies on muqarnas corbelling mask these and other transitions. Such balconies inevitably recall those of Ottoman minarets, and indeed were used to secure the same antiphonal effects in the chanting of the adhan as in Turkey. Built into the crowning cupola were a series of projecting poles from which lamps were sus- 122 Cairo, Mughalbay Taz mosque, minaret 3.129 3.125 170
3.4, 3.123 c f 3.113, 3.120 3.126 2.34 c f 2.33 1119-, cf. 3.130 pended on the occasion of the great festivals (minaret of Bashtak, 737/1336). Thus the ancient associations built into the very name of the structure were perpetuated. A new emphasis on absolute height may be discerned in the minarets of the later Mamluk period, such as that placed at the south-east corner of the Sultan Hasan mosque (757-60/ 1356-9), which soars to 280 feet, and is the tallest in Cairo. This example is also typical of the later period in that the crowning dome is carried on an open circular colonnade - a tholos, in fact, allotted a new and quite unexpected function (minaret of Aqsunqur, 748/1347). Sometimes these columns are doubled. The crowning element of the minaret also changes definitively under the rule of the Burji Mamluks, though the first examples of the new form date from the early 14th century. Earlier, the crowning feature was the diminutive two-storey mabkhara - so called because it resembled the top of an incense-burner, though they are also locally known as pepper-pots. Now this was replaced by the qulla, which owed its name to its resemblance to the upper half of the typical Egyptian water-container. The pear-shaped qulla usually bears at least two bronze finials whose crescents are orientated towards the qibla. In the final decades of Mamluk rule a playful variation on this theme makes its appearance: the minaret is crowned by a pair of pavilions, square in plan and crowned by a whole cluster of qullas (funerary complex of al-ghuri; one might compare the minaret of al-ghuri in the Azhar mosque and the minaret of the Qani Bay mosque). It is entirely fitting that the evolution of the medieval Egyptian minaret should end on this fanciful note, for the previous five centuries had shown lavish decoration to be the keynote of this tradition. The changing succession of geometrical forms - principally cube, rectangle, octagon and cylinder - allowed free rein to this decorative emphasis, which is unmatched in any other group of minarets. Finally, the popularity of the minaret in Mamluk architecture invites explanation. In the 14th and 15th centuries the main building type in Cairo appears to have been the composite ensemble. Its constituent parts could vary from one ensemble to another, but their main functional elements were the mosque, madrasa, khanqah and mausoleum. Similar complexes had already become popular in Saljuq Anatolia. In Egypt, however, unlike Anatolia, the minaret was from the first regarded as an integral part of such complexes. Whether this was entirely for functional reasons may be doubted. In the dense urban fabric of Cairo nothing could more appropriately designate such a complex from afar than a minaret. In this sense it could be regarded as a public affirmation of its patron s munificence, and thus served a personal, quasi-totemic function. Their placing varied. Sometimes they were located at the two corners of the principal fa ade, or flanking a gateway (e.g. Bab Zuwaila); these were traditional locations. But many of the locations were unusual or even unprecedented. The madrasa of al-salih has a single minaret above the central porch of the fa ade, and the two minarets in the mosque of al-nasir Muhammad on the citadel are at one corner of the qibla wall and to one side of the main entrance. The latter location recurs in the funerary complex of Qa it Bay. In this unpredictable positioning of the minaret one may recognise concerns similar to those of Ottoman architects. Now the minaret was, it seems, valued less for its actual or symbolic religious function and more for its role as a marker or articulating feature, both within the complex to which it belonged and, more broadly, within the cityscape itself. Once again, the flexibility of the forms developed by Islamic architects had asserted itself. 3.124 3.131 171